Data-driven network analysis methods provide new perspectives for the understanding of emission reduction synergy. This paper sorts out the historical logic, spatial characteristics and driving factors of China-style modernization regional emission reduction policy synergy from both historical and practical perspectives, in order to provide reference for policy formulation. Different from the existing investigation of the synergy between government departments, this paper constructs an inter-provincial policy network measurement model based on natural language processing, based on big data text analysis to measure the coordination degree of 4988 inter-provincial and regional policies and 275 central policy concerns, introduces social network analysis technology to construct an inter-provincial policy coordination network, and studies the impact of central policy focus on local emission reduction policy coordination. Based on the BERTopic model, it is found that the regional coordinated emission reduction policy has gone through three stages: radiation in key demonstration areas, regional coordinated strategic guidance and national systematic classification policy, and three mechanisms have been formed: differentiated target response, expanded synergy scope and concrete policy content. Applied social network analysis shows that there are 457 emission reduction synergies coexisting among 31 provinces, and the network shows a tight imbalance between the central and eastern parts and the loose imbalance in the western and northern parts of the south. Among them, Beijing, Shanghai and Jiangsu are the core actors, Jilin, Ningxia and Heilongjiang are the marginal actors, and Hubei and Hunan play the role of "bridges"; Finally, based on the secondary assignment procedure of multiple regression, it is found that the higher the attention of the central government, the closer the geographical location, and the greater the gap in economic development level, the more conducive to regional emission reduction coordination. The central government should balance policy concerns, adhere to quantitative targets and structural optimization, and implement policies according to role positioning.

数据驱动下的网络分析方法为减排协同的认知提供了新的视角。本文从历史和实际两个角度梳理中国式现代化区域减排政策协同的历史逻辑、空间特征和驱动因素,以期为政策制定提供参考。与既有考察政府部门间的协同关系不同,本文构建基于自然语言处理的省际政策网络度量模型,以大数据文本分析度量4988份省际区域政策协同度以及275份中央政策关注度为基础,引入社会网络分析技术构建省际政策协同网络,研究中央政策侧重对地方减排政策协同的影响。基于BERTopic模型发现区域协同减排政策经历了重点示范地区辐射、区域协调战略引领和全国系统分类施策三个阶段,形成了差异化的目标回应、扩大化的协同范围以及具体化的政策内容三种机制;应用社会网络分析发现31省份间共存在457次减排协同关系,网络呈现出中部、东部紧密而西部、南北部松散的不平衡态势。其中,北京、上海、江苏是核心行动者,吉林、宁夏、黑龙江是边缘行动者,湖北、湖南起到 “桥梁”作用;此外,基于多元回归二次指派程序发现中央政府关注度越高、地理区位越近、经济发展水平差距越大越有利于区域减排协同。中央政府应平衡政策关注、坚持数量目标和结构优化、依据角色定位分类施策。